Figure 1

Methodological workflow. After Whole Exome Sequencing analysis of 27 CMM families, the resulting variants were filtered, and methods based on the network propagation principle were applied to prioritize candidate genes in the vicinity of genes previously related to CMM. Seed genes are known CMM genes (including susceptibility genes and somatic drivers) and candidate genes are those identified from the whole-exome sequencing analysis. The network propagation amplifies a biological signal based on the assumption that genes underlying similar phenotypes tend to interact with one another. Each gene was scored by its similarity to every other gene in protein interaction networks (interactomes). These scores or probabilities were then used to rank candidate genes and reveal gene clusters, respectively. Degree aware algorithm (DADA) was applied for gene ranking, and Hierarchical HotNet and GeneMANIA tools were used to identify modules. The variant filtering plan excluded variants based on population frequency in databases and internal controls, predicted pathogenicity, cosegregation in families, and quality control measures (see “Methods” section for filtering details).